In situ combustion is a possible method for producing heavy oil when other methods such as SAGD are not adequate (e.g., in thin beds, or when CO2 emission for steam generation is unacceptable). Previous field trials of this process have often been unsuccessful. However, in recent years, several new well implementations have been proposed (COSH, THAI), exploiting the more advanced drilling capabilities now available. Simulations of such configurations require a reliable representation at field scale of the oxy-combustion reactions, which is not available at the present time. The objective of the work described in this paper is to illustrate some improvements in the description of oxy-combustion reactions both at the experimental level and in the simulation models. A "ramped temperature" experiment has been conducted on an extra heavy stock tank oil (10,000 cP at reservoir conditions). This experiment has been successfully matched using a commercial simulator. The improvement over classical adiabatic reactor experiments is significant: two combustion reactions are clearly observed, and the Arrhenius parameters are determined with increased accuracy. The reliability of the inferred parameter values is checked by applying them to simulations of previous adiabatic disk reactor experiments conducted under a variety of conditions. The final part of the paper is dedicated to illustrating the impact of the new reaction scheme on the simulation results at field scale. Introduction With the more advanced drilling capabilities now available, such as horizontal wells, several new well configurations have recently been proposed for in situ combustion applied to heavy oil [COSH(1), THAI(2)]. To assess the potential of these new configurations by simulation, a reliable representation of the oxy-combustion reactions is required. These reactions govern the oxygen consumption and the time of oxygen breakthrough. Consequently, they will directly influence the efficiency of the recovery process in any given well configuration. Previous authors have addressed the topic of oxy-combustion reactions and kinetics, and a number of publications are available in the literature(3–15). One of the major literature contributors is the University of Calgary, notably due to their work in determining/ proving the presence of the high/low temperature oxidation zones(3, 4). On our side, in order to improve the description of the oxy-combustion reaction scheme both at the experimental level and at the numerical simulation level, we have conducted and simulated a new type of ramped temperature experiment with an extra heavy oil. This new type of experiment has been successfully matched on a commercial simulator (STARSTM from CMG) and has led us to develop a new reaction scheme with two combustion reactions. In the final part of the paper, the impact of this new reaction scheme is evaluated at field scale by comparing numerical simulations made with the old and with the new reaction schemes. Ramped Temperature Experiment Motivations Before developing the ramped temperature experiment, two types of experiments had been developed in Total's Thermal Methods Laboratory in order to determine the kinetic parameters of the oxy-combustion reactions for light oils(9):

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractThis paper describes a new technique to decrease the computational times of thermal simulations. Effectively, thermal processes are based on the displacement of a thermal front (combustion front, steam chamber interface), around which most fluid flows will take place. Thus, we propose a dynamic gridding approach, to keep a fine scale representation around the thermal front, and a coarser grid away from the front, thus leading to cheaper computations.We will first describe the principles of this dynamic gridding. Simulations will start with an original fine grid, but will reamalgamate its cells, while keeping some regions (for example around wells) always finely gridded. The gridding will then identify the moving front through large gradients of specific properties (temperatures, fluid saturations and compositions). In the front vicinity, it will de-amalgamate the originally amalgamated cells, and later on re-amalgamate them once the front has passed. Amalgamated cells are assigned up-scaled properties, this upscaling being based upon classical averaging techniques.We will illustrate this dynamic gridding technique with simulation examples, as it has been successfully implemented in a thermal simulator, STARS, a product of Computer Modelling Group Ltd (CMG). Using examples on combustion and SAGD simulations, we will show that it can divide the CPU time of thermal simulations by a factor of 2 to 3, without loss of accuracy.